Nahuatl Dictionaries Online: The Good, the Bad and the Really Terrible.

There is no shortage of dictionaries of Nahuatl. Throughout the colonial period and especially since the 1940s dozens of small dictionaries and wordlists of Nahuatl have been published. But most of them are not available online. 

In this blog post I am going to review some of the Nahuatl dictionaries that can be found online. Some of them are really excellent, good enough to be used by scholars and academics, others are so terrible that they shouldn't be used by anyone at all. 

First let me point out that there is wide agreement among academics that when it comes to Nahuatl dictionaries, the gold standard is Alonso Molinas 1571 "Vocabulario en lengua castellana y mexicana (1571)". This work is unfortunately not available online, but can be bought in facisimile from different Mexican book traders. This is a must have for any Nahuatl scholar. [Edit: In the comment section Ayac has noted that it IS available online in facisimile so here I add the link https://archive.org/details/vocabularioenlen00moli  The internet truly is beautiful.]

Newer, well esteemed, print dictionaries include Frances Karttunen's Analytical Dictionary of Nahuatl (cover pictured on the right), which is particularly useful for scholars because it includes vowel length and glottal stops. It is based on Molina's vocabulario n and Carochi's grammar and includes information from a couple of dictionaries of modern varieties (especially it uses evidence from tetelcingo Nahuatl to establish vowel length when Carochi does not record it). Many scholars use Remi Simeon's 1885 dictionary which has been published in many subsequent editions. I personally don't use it both because even though it is very big and has more entries than most other dictionaries, it is a derivative work and its sources are unclear and difficult to trace. This means that it is hard to check for errors, and to trace possible dialectal or regional variation in vocabulary. And it does contain a lot of errors as derivative works inevitably do (both the errors of the sources and their own new errors). But mostly I don't use it because it is brick-sized, expensive and there are better dictionaries online. Here are some of them: 

The Good:

Wired Humanities Dictionary: http://whp.uoregon.edu/dictionaries/nahuatl/

This dictionary put online by the Wired Humanities project of the University of Oregon, directed I believe by Nahuatl ethnohistorian Stephanie Wood, really does everything you want of a Nahuatl dictionary. It translates both ways to both Spanish and English, it includes vowel length and glottal stops (as well as a not very useful phonetic transcription of each entry), it gives examples of usage from actual texts and it is easy to search. It has a really good number of entries although on more than one occasion, I have drawn a blank on some odd word. The sources are clearly marked, it separately gives the definitions verbatim from each of them. The sources include Molina, Karttunen, and sometimes it includes entries from the IDIEZ monolingual dictionary of Modern Huasteca Nahuatl which has not yet been published. Sometimes it gives too many entries, especially on short words, because it returns all the words that contain a certain string of letters. Then you have to scroll down to the one you want. 

Alexis Wimmer's Dictionnaire de la langue Nahuatl classique: http://sites.estvideo.net/malinal/nahuatl.page.html
Alexis Wimmer's dictionary is basically an electronic version of Simeon's dictionary all of the material in which it includes. Except it is better. What makes it better is that it includes examples of usage of all the entries, taken from the Florentine Codex. Sometimes there are many examples and sometimes few, but they are always an immense help in understanding how a word is used and its deeper contextual meanings. The main drawbacks of this dictionary is that it is not searchable, you have to enter by first letter, then find the interval of letters where the word is located and then search visually (or with the browser's search function) for the entry you are looking for. This means that it is basically just as slow to find a word as if it were a paper dictionary. The second drawback is that it is in French, sometimes with Spanish translations in the examples. This means that sometimes I have to get out a french dictionary or do a webtranslation for the french word if I don't know it. Nonetheless this has become my general go-to dictionary because it has so many entries combined with the great examples. 

The Not So Good:  

SIL dictionaries: http://www-01.sil.org/mexico/pub/dicgram.htm#FamiNahuatl
Through their Mexican branch the Instituto Linguistico de Verano who had a kind of monopoly on making indigenous language educational materials from the 1930s to 1980s, the missionary organization the Summer Institute of Linguistics made a lot of Nahuatl dictionaries of specific varieties. Many of them are available for download at their website linked below. The good thing about these is that they are each based on fieldwork with a single variety and reflect local usage form the communities studied. These makes them pretty much the only current source of lexical knowledge about several Nahuan varieties. They are often written by two authors with the same surnames - that is because many of them are missionary couples who lived in the community for decades working simultaneously on the dictionary and Bible translations into the language. Some of them had little linguistic training, others are excellent linguists. The early ILV dictionaries were published in a series called Mariano Silva y Aceves, these are the worst ones as they are all very short give only a single word meaning for each entry and no examples - basically just wordlists. Some of them though like the Brewers' dictionary of Tetelcingo Nahuatl, and Wolgemuth's dictionary of Mecayapan Nahuat  is still useful because there are no other ones for this highly divergent variety. The more recent ones are made much more professionally with examples and grammatical information. Some of them, like the Brockways' otherwise excellent dictionary of Nahuatl from the Sierra de Puebla are marred by odd lexicographic choices such as listing transitive verbs with the object prefix attached (meaning that they are all listed under qui-, tla- or mo-). These are not really online dictionaries though, but just paper dictionaries put online.

CONDIP: http://www.condip.dk/condip.php
It breaks my heart having to list the COpenhagen Nahuatl DIctionary Project CONDIP, in the not so good section. Both because it comes out of my Alma Mater the University of Copenhagen and is the work of my mentor Una Canger and also because it really should be in the good section, it has just been superseded very quickly by technology. It is essentially an online dictionary 1.0, when today we live in a 4G world. Canger very early on saw the promise of digital technology for lexicography and the linguistic need for dictionaries that could use several orthographies and link words directly to a text corpus. This is what CONDIP was designed to be by Canger and the programmer who coded it Michael Thomsen. It uses three orthographies, a phonemic orthography with symbols based on the Americanist Phonetic Alphabet, a standardized classical orthography and the actual paleography of the texts - this is really useful for linguistic analysis. And it also gives morphological information in the phonemic representation which also represents the morphological analysis. For example in text view the word for Coyote may be written "coiotl" as it is written in the Florentine Codex,  but in analytical view it is "koyō+λ"(the + is a morpheme boundary).

But due to the era in which it was designed, the early 1990s, the technology platform was DOS based (requiring basic coding skills to use) and there was no user Graphic User Interface. We didnt think of this as a problem back then in the 1990s of course, but today it just doesnt work very well with how we use technology. Secondly the text data base had to be handcoded and hence turned out to be quite small compared to how big one expects a database to be today. It is based on Books 2 and 11 of the Florentine codex, the codex Aubin, parts of Molina (perhaps all if more coding has been accomplished since I was last informed) the Bankcroft dialogues and a couple of colonail documents. It really can be a very good tool to find the text examples, but the Wired Humanities dictionary mentioned above just is much better in terms of the user interface, and the demands put on the user. I was still very happy when I saw that the dictionary had been put online and works as an actual online dictionary. But except for a couple of time when I tried it out of nostalgia I really haven't used it - instead I use the two ones listed in good. Canger published a really good article about the dictionary in the 2002 volume "Making Dictionaries" edited by William Frawley, Kenneth Hill and Pamela Munro. This one is worth a read just to read the way Canger describes all the challenges and pitfalls of creating a dictionary of a language like Nahuatl - the only analytical description of such problems that I know of. 

Xochitlahtolli Zongolica Nahuatl Dictionary:  http://www.vcn.bc.ca/prisons/dicc-zon.pdf

Another really decent dictionary available online is the one made by the Veracruzan collective of Nahuatlahtos Xochitlahtolli. This group of native speaking scholars have produced this valuable dictionary of the varieties of the Zongolica region in collaboration with linguist Andres Hasler. It is not huge, but have a really decent amount of entries, and is particularly interesting because it describes a regional variety (with some attention to variation between communities in the region) and is written mainly by native speakers. It does not include vowel length even though this is an important feature of many varieties of the region - but this is probably because this is less important for native speakers who produce the right vowel length intuitively. It is a searchable pdf file, but basically a print dictionary put online. It is also interesting because it uses an orthography that is not based on the classical system but which uses k and w, instead of the classical qui/que, hui/hua etc.



The Really Terrible:

AULEX: http://aulex.org/es-nah/?idioma=en
Generally it is a bad idea to plug bad webresources since it is usually better to just let them go by in silence. But I am including this dictionary here because I have already met several people online who have used it, and I feel responsible for issuing a general warning about it. Aulex is a kind of "crowd sourced" dictionary which means that it is basically made by ordinary folks who compile the dictionary and upload it to the website. The Nahuatl-Spanish/Spanish Nahuatl version of this dictionary is made by Manuel Rodriguez Villegas. In the prologue he gives a description of the process of making it, and an outline of Nahuan varieties, and a list of the sources he has used. The outline of the dialects is an entirely erroneous and confused list of different Nahuatl varieties. He distinguishes "Nahuatl Moderno" (Modern Nahuatl) as if it were a separate variety (it is a cover term for all the currently existing varieties); and randomly mentions a couple of specific local varieties from regions or specific communities, as if they were on the same level; he uses idiosyncratic neologistic names for the varieties based on various ethnohistorical sources that he does not list. He also invents the absurd language "Nahuatl Paiute-Apache" which is non-existent and seems to be a combination of Paiute a Numic language that is distantly related to Nahuatl and Apache which belongs to the Athabascan language family and has nothing in common with Nahuatl at all.  He also claims Nahuatl to be spoken in several places where it is not. 

The worst part however is the sources, which is an eclectic mix of other good dictionaries, bad Spanish language paper dictionaries of specific dialects and terrible pseudolinguistic works such as the works by Juan Luna Cardenas who made up new Nahuatl words for all kinds of things and put them in books. For example he famously coined the word "ixachilan" which he alleged to refer to the American continent, but which in reality has never been used by actual native Nahuatl speakers. In the dictionary this word however is randomly changed to ixachitlan, which is not what Luna Cardenas actually wrote. Tonnes of entries have these kinds of errors, giving for example sometimes word forms that are phonologically impossible in any known variety of Nahuatl. It also contains hundreds of what seem to be newly coined neologisms for modern concepts, such as US states, (Alaskak "Alaska", kaliforniatl "person from California"), or technologies (pipiyolkali "helicopter"). Most of these are unintelligible to native speakers (pipiyolkalli forexample would seem to be composed of the roots for "bee" and "house", giving the logical meaning "beehouse" or "hive", not "helicopter"). This means that trying to pass of anything written with the dictionary as actual Nahuatl would be confusing and possibly offensive to native speakers who have no chance of knowing these words or their meanings.  

The dictionary never gives examples of usage, and it is impossible to trace the source of a given entry. This is particularly problematic when the sources come from individual local varieties that have different phonological systems so that a word in one dialect is impossible and nonsensical in another - such as when the dictionary gives words with the vowel /u/ when in the vast majority of Nahuatl varieties do not have this vowel. Needless to say it doesnt record glottal stops or vowel length. 

No one should use this dictionary for any serious purpose, as it is an insult both to scholarship and to the people who speak the language natively.